Theories of protection motivation imply that one enduring behavioral consequence of trauma should be an increase in precaution. The applicant's study of psychosocial functioning following Hurricane Hugo has created an opportunity to examine the effects of disaster specifically and trauma more generally on victims' efforts to protect themselves from subsequent traumatic events. This study would address four major questions: (l) What is the frequency of precautionary behavior? Behaviors related to hazard preparedness crime prevention vehicular safety, and health maintenance would be described. (2) What is the structure of precautionary behavior? Hypothesized measurement models relating these behaviors to underlying strategies and domains would be confirmed. (3) What are the precursors of precautionary behavior? The additive contributions of prior disaster or trauma, various personal and situational constraints, and perceived controllability of events would be assessed. Of special interest is whether event-induced change is specific to the domain of concern or whether it generalizes to other precautionary domains. And (4) What are the consequences of precautionary behavior? Does precaution lessen one's probability of experiencing traumatic events, does it lessen the trauma potential of events that do occur, or does it merely augment one's sense, or illusion, of personal control? The design of for the proposed study is longitudinal. An interview conducted in October 90 provided information on exposure to Hugo and other events, such as physical assault, motor vehicle accidents, and fire.Exposure to disaster was assessed at both individual and community levels. An interview to hazard preparedness, crime prevention, vehicular safety, and health maintenance. An interview to be conducted in October 9l provided information on behaviors and attitudes toward behaviors related to hazard preparedness, crime prevention, vehicular safety, and health maintenance. An interview to be conducted in October 94 would provide information on events and reactions to events experienced over the 5-year interval since Hurricane Hugo. That these three timepoints serve to separate the measurements of precursors, behaviors, and consequences is a strong methodological advantage. The October 94 sample should number 650-700 and should be approximately half female and half Black.